… High-dimensional and sparsematrices (HiDS) matrices with non-negativity constraints are … sparsenon-negativematrix factorization model for largescalerecommendersystems on GPUs…
M Li, L Sheng, Y Song, J Song - Expert Systems with Applications, 2022 - Elsevier
… In many industrial applications related to big data, a so-called high-dimensional sparsematrix is often used to represent relationships between entities, where the sparsity mainly results …
X Shi, Q He, X Luo, Y Bai… - IEEE Transactions on Big …, 2020 - ieeexplore.ieee.org
… Li, “Efficient extraction of non-negativelatentfactors from high-dimensional and sparse matrices … Zhu, “A nonegative latentfactormodel for large-scalesparsematrices in recommender …
J Chen, X Luo, MC Zhou - … on Human-Machine Systems …, 2020 - ieeexplore.ieee.org
… Therefore, more and more researchers have shown their interests in Recommender Systems (RSs) [1-5], which play an important role in assisting people to obtain their needed …
W Li, Q He, X Luo, Z Wang - IEEE Transactions on Systems …, 2020 - ieeexplore.ieee.org
… of the proposed SNbased model in building CF-based recommendersystems, its accuracy … A Nonnegativelatentfactormodel for large-Scalesparsematrices in recommendersystems …
J Jiang, W Li, A Dong, Q Gou, X Luo - Neurocomputing, 2020 - Elsevier
… matrix between the input layer and the hidden layer. Thus, in an FAE-based model, we can map the input vector r i to the latentfactor … Note that for a large-scale optimization problem, …
G Behera, N Nain - International Journal of Information Technology, 2022 - Springer
… non-negativematrix factorization (DNNMF) technique to address the above problem. First, we impose non-negative … We design a nonlinear non-negativematrix factorization using deep …
J Chen, Y Yuan, T Ruan, J Chen, X Luo - Neurocomputing, 2021 - Elsevier
… However, an LFA model relies heavily on its hyper… latentfactor analysis (HLFA) model. Its main idea is to build a swarm by taking the hyper-parameters of every single LFA-based model …
… application in large-scalerecommendersystems, the … In recommendersystems, because many users have sparse … uk denotes the preference of user u on the latentfactor k ( k ∈ { 1 , 2 , …